Authors:Marwan Ramdhany Edy, S. Wahjuni, S. N. Neyman Pages: 145 - 152 Abstract: Some jobs require that the concentration level be maintained for a long time during work time. Lack of sleep will in disruption of someone concentration level. To find out a human concentration level can be done by recording his/her brain waves. This research uses Electroencephalography (EEG) technology which functions to capture human brain waves. The focus of this study is to build a model of the detection system of a human concentration level. The research datasets are data from brain wave recording using Neurosky Mindwave Mobile which has extracted in 19 features. Data will then be labeled using cluster techniques namely Fuzzy C-Means to become data to be input into the classification process using Support Vector Machine (SVM). The classification results show an accuracy of 98.34%. That results show FCM can be used to automatically label EEG data properly. PubDate: 2019-09-24 DOI: 10.18495/comengapp.v8i3.307 Issue No:Vol. 8, No. 3 (2019)
Authors:Kenga Mosoti Derdus, Vincent Omwenga Oteke, Patrick Job Ogao Pages: 153 - 158 Abstract: Virtualization has enabled cloud computing to deliver computing capabilities using limited computer hardware. Server virtualization provides capabilities to run multiple virtual machines (VMs) independently in a shared host leading to efficient utilization of server resources. Unfortunately, VMs experience interference from each other as a result of sharing common hardware. The performance interference arises from VMs having to compete for the hypervisor capacity and as a result of resource contention, which happens when resource demands exceed the allocated resources. From this viewpoint, any VM allocation policy needs to take into account VM performance interference before VM placement. Therefore, understanding how to measure performance interference is crucial. In this paper, we propose a simple experimental approach that can be used to measure performance interference in Infrastructure as a Service (IaaS) cloud during VM consolidation. PubDate: 2019-09-24 DOI: 10.18495/comengapp.v8i3.311 Issue No:Vol. 8, No. 3 (2019)
Authors:Olalekan Ogunbiyi, Cornelius T Thomas Pages: 159 - 172 Abstract: Nigeria with a generating potential of roughly 12,522 MW only supplies less than 20% of the national demand. This necessitates an optimal use of the Jebba Hydroelectric Power Plant whose optimal generation depends on the operating head. This paper presents the solution to an optimal control problem involving the operating head of the plant. An optimal control problem consisting of a model of the system dynamics, performance index and system constraints was solved using a dynamic programming approach. The control procedure was built on the integration of the nonlinear dynamical model by an Adams-Moulton technique with Adams-Bashfort as predator and Runge-Kutta as a starter. The numerical solution, coupled with dynamic programming was employed in developing an optimal control procedure for the regulation of the operating head. Result presented shows the potential of the control procedure in determining the amount of inflow required to restore the operating head to a nominal level whenever there is a disturbance. PubDate: 2019-09-24 DOI: 10.18495/comengapp.v8i3.315 Issue No:Vol. 8, No. 3 (2019)
Authors:Benjamin Kommey, Seth Djanie Kotey, Andrew Selasi Agbemenu Pages: 173 - 179 Abstract: A number of accidents on our roads are caused by driver fatigue or drowsiness. Human fatalities as a result of driver drowsiness has been a major challenge for road safety bodies worldwide. Various road safety campaign messages have been put out to discourage drivers from driving whilst tired, but the problem still persists. Different technologies have been proposed over the years, but most seem to be too expensive to implement on a large scale. We present an inexpensive drowsing driver alert system in this paper. The system, known as Drowsing Driver Alert System (DDAS) is a smart system intended to effectively keep commercial drivers alert when driving. The system is able to detect when a driver is drowsy and alert him/her in real-time to prevent a potential accident. Using a camera, the eyes of the driver are monitored continuously whiles driving and analyzed to determine if they are shut or the blink rate is not normal. Two stages of alerts are given if the driver is determined to be drowsy. Log files of activities performed by the system are also saved to an external storage device to enable further analysis later. PubDate: 2019-09-24 DOI: 10.18495/comengapp.v8i3.308 Issue No:Vol. 8, No. 3 (2019)
Authors:Nyayu Husni Latifah, Masayu Annisah, Tresna Dewi Pages: 181 - 188 Abstract: In this research, a fuzzy logic algorithm is implemented in a monitoring system for detecting the potential fires in peat land. The monitoring system in this research employs two sensors as the fuzzy inputs, i.e. TGS 2600 gas sensor and DHT11 temperature sensor. The outputs of the fuzzy logic are the specified conditions of motor activation (PWM) with 3600 rotations. The system is monitored through camera, which sends the monitoring result to android via web server. The result is sent when TGS 2600 and DHT11 sensors detect the determined gas concentration and surrounding temperature. Before sending the result, the rotating motor stops every five minutes to take the photograph of peat land location. The result shows that the algorithm used in this research has been successful in determining the condition of the peat lands correctly and therefore can be used as the early prevention of fires. PubDate: 2019-09-24 DOI: 10.18495/comengapp.v8i3.297 Issue No:Vol. 8, No. 3 (2019)
Authors:Nurul Afifah, Deris Stiawan Pages: 189 - 202 Abstract: Malware is very dangerous while attacked a device system. The device that can be attacked by malware is a Mobile Phone such an Android. Antivirus in the Android device is able to detect malware that has existed but antivirus has not been able to detect new malware that attacks an Android device. In this issue, malware detection techniques are needed that can grouping the files between malware or non-malware (benign) to improve the security system of Android devices. Deep Learning is the proposed method for solving problems in malware detection techniques. Deep Learning algorithm such as Deep Neural Network has succeeded in resolving the malware problem by producing an accuracy rate of 99.42%, precision level 99% and recall 99.4%. PubDate: 2019-09-24 DOI: 10.18495/comengapp.v8i3.294 Issue No:Vol. 8, No. 3 (2019)
Authors:Renny Amalia Pratiwi, Siti Nurmaini, Dian Palupi Rini Pages: 203 - 216 Abstract: Melanoma causes the majority of skin cancer deaths. The population level of melanoma has increased over the past 30 years. It kills around 9.320 people in the US every year. Melanoma can often be found early, when it is most likely to be cured. Medical diagnoses using digital imaging with machine learning methods have become popular because of their ability to recognize patterns in digital images. Image diagnosis accuracy allows disease cured at an early stage. This paper proposes a simulation that can be used for early detection of skin cancer that can help dermatologists to distinguish melanomas from other pigmented lesions on the skin. Some researchers have developed a system using machine learning algorithms used to classify skin lesions from dermoscopy images of human skin. In this study, we proposed Convolutional Neural Network (CNN) to our model. CNN is very efficient for image processing because feature extractors can be optimized, applied to each feature image position. The results of skin lesion classification of benign nevi and melanoma based on CNN models produces high accuracy (area under the receiver operator characteristics (ROC) curve (AUC) is 92.59 %, sensitivity is 89.47%, specificity is 100.0%, precision is 100 % and F1 score is 94.44 %). PubDate: 2019-09-24 DOI: 10.18495/comengapp.v8i3.290 Issue No:Vol. 8, No. 3 (2019)
Authors:Xiaojun Ma, Shengjun Zhai, Yingxian Fu, Leonard Yoonjae Lee, Jingxuan Shen Pages: 217 - 227 Abstract: This paper looks at the problem of employee turnover, which has considerable influence on organizational productivity and healthy working environments. Using a publicly available dataset, key factors capable of predicting employee churn are identified. Six machine learning algorithms including decision trees, random forests, naïve Bayes and multi-layer perceptron are used to predict employees who are prone to churn. A good level of predictive accuracy is observed, and a comparison is made with previous findings. It is found that while the simplest correlation and regression tree (CART) algorithm gives the best accuracy or F1-score, the alternating decision tree (ADT) gives the best area under the ROC curve. Rules extracted in the if-then form enable successful identification of the probable causes of churning. PubDate: 2019-09-24 DOI: 10.18495/comengapp.v8i3.316 Issue No:Vol. 8, No. 3 (2019)
Open Access journal
[4 journals]
